Authentication method and apparatus, and method and apparatus for training a recognizer
First Claim
Patent Images
1. An authentication method comprising:
- applying an input image to a recognizer, the recognizer including a plurality of layers of a deep learning (DL) network, the input image including a first object; and
determining whether the first object is the same as a second object in a registered image, the determining including,determining a similarity or a distance between at least one feature value of the input image and at least one feature value of the registered image corresponding to a first layer of the layers in the recognizer, the at least one feature value of the input image output by the first layer of the layers in the recognizer,determining whether the determined similarity or distance is within a threshold range corresponding to the first layer of the layers in the recognizer, the threshold range corresponding to the first layer of the layers in the recognizer being different than a threshold range corresponding to a second layer of the layers in the recognizer, andauthenticating the input image based on whether the similarity or the distance is within the threshold range corresponding to the first layer of the layers in the recognizer.
1 Assignment
0 Petitions
Accused Products
Abstract
An authentication method and apparatus, and a method and apparatus for training a recognizer are provided. The authentication method may include determining whether authentication of an input image succeeds by comparing a similarity associated with an intermediate layer to a threshold.
40 Citations
35 Claims
-
1. An authentication method comprising:
-
applying an input image to a recognizer, the recognizer including a plurality of layers of a deep learning (DL) network, the input image including a first object; and determining whether the first object is the same as a second object in a registered image, the determining including, determining a similarity or a distance between at least one feature value of the input image and at least one feature value of the registered image corresponding to a first layer of the layers in the recognizer, the at least one feature value of the input image output by the first layer of the layers in the recognizer, determining whether the determined similarity or distance is within a threshold range corresponding to the first layer of the layers in the recognizer, the threshold range corresponding to the first layer of the layers in the recognizer being different than a threshold range corresponding to a second layer of the layers in the recognizer, and authenticating the input image based on whether the similarity or the distance is within the threshold range corresponding to the first layer of the layers in the recognizer. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20)
-
-
21. A training method comprising:
-
training a recognizer, the recognizer including a plurality of layers of a deep learning (DL) network; determining a first threshold range for a first layer in the trained recognizer based on a desired performance index, the first threshold range being used for early authentication of an input image, the early authentication being a termination of an authentication method before reaching a final layer of the plurality of layers in the trained recognizer based on the first threshold range and authentication of the input image; and determining a second threshold range for a second layer in the trained recognizer based on the desired performance index. - View Dependent Claims (22, 23, 24, 25, 26, 27, 28)
-
-
29. An authentication apparatus comprising:
at least one processor configured to execute computer-readable instructions to cause the authentication apparatus to, output at least one feature value of a first object in an input image using a first layer among a plurality of layers of a deep learning (DL) network included in a recognizer, and determine whether the first object is the same as a second object in a registered image by determining a similarity or a distance between the at least one feature value of the input image output using the first layer and a feature value of the registered image corresponding to the first layer, determining whether the determined similarity or distance is within a threshold range corresponding to the first layer of the layers in the recognizer, the threshold range corresponding to the first layer of the layers in the recognizer being different than a threshold range corresponding to a second layer of the layers in the recognizer, and authenticating the input image based on whether the similarity or the distance is within the threshold range corresponding to the first layer of the layers in the recognizer. - View Dependent Claims (30, 31, 32, 33, 34, 35)
Specification